Real-time systems are often difficult to analyze. Such is the case for both hardware and software based systems. In hardware based systems, for example, an analyst may not be able to access a desired signal, much less, access the desired signal and causes of such signal in real-time. In fact, in some systems, even the mere act of accessing an available signal may have detrimental effects on the system operation.
In software based systems, for example, standard analysis techniques such as setting breakpoints in a debugging environment and stepping through a software program do not typically work well for real-time systems. For example, once the software program execution is stopped in the debugger, the relationship between the program's execution and the real-time chain of events is broken. Events continue to occur in real-time while the program is executed at a completely different non-real-time rate under the control of the debugger. Such single stepping through a program may not reveal the performance information sought since, by the time the section of software code of interest is executed, the conditions causing the event of interest have passed.
Also in software-based systems, for example, engineers have traditionally relied on debug messages in the program that normally include text strings and indicate which events are taking place, which piece of code is executing, and the values of certain parameters. However, such a technique does not provide a complete performance analysis solution since such messages cannot describe everything happening in the system. Using this traditional technique, debug messages would likely have to be incorporated into every part of the program where execution flow changes (e.g., in every “if” and “while” statement). The amount of debug messages may be excessive in such a case, and in fact, the execution of such debug messages may interfere with real-time system timing.
Further, when input data to a real-time system (e.g., a communication system) changes from one operation to the next, operational analysis becomes increasingly difficult. An analyst may need to perform a multitude of test trials before a problem can be reproduced, analyzed, understood and fixed. Another complication that may occur (e.g., in communication systems) is when the problem occurs only at a remote site due to the variance of the operational conditions. A trip to the remote site by one or more engineers may be required and can be time consuming and expensive.
Further limitations and disadvantages of conventional and traditional approaches will become apparent to one of skill in the art, through comparison of such systems with the present invention as set forth in the remainder of the present application with reference to the drawings.